JOURNAL OF MORPHOLOGY 00:1–10 (2014)

Ontogenetic Development and of Franciscana Skull: A 3D Geometric Morphometric Approach

Daniela L. del Castillo,1,2* David A. Flores,2,3 and Humberto L. Cappozzo1,2,4

1Laboratorio de Ecologıa, Comportamiento y Mamıferos Marinos, Division Mastozoologıa, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia” Av, Angel Gallardo 470 (C1405DJR), Buenos Aires, 2CONICET, Consejo Nacional de Investigaciones Cientıficas y Tecnicas, Buenos Aires, Argentina 3Division Mastozoologıa, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia” Av, Angel Gallardo 470 (C1405DJR), Buenos Aires, Argentina 4Departamento de Ciencias Naturales y Antropologıa, Centro de Estudios Biomedicos, Biotecnologicos, Ambientales y Diagnostico, Fundacion Azara, Universidad Maimonides, Hidalgo 775 7mo (C1405BDB), Buenos Aires, Argentina

ABSTRACT The aim of this work was to study the Odontocete cetaceans follow the common pattern postnatal ontogenetic development of Pontoporia blain- of , that is, sexual dimorphism with villei skull, identifying major changes on shape, and increasing body size (Ralls, 1977; Tolley et al., relating them to relevant factors in the life history of 1995). However, a few odontocete species show a the species. We analyzed a complete ontogenetic series less common pattern of inverted sexual dimor- (73#,83$) with three-dimensional geometric morpho- metric techniques. Immature showed a very phism in size, and females usually reach larger well-developed braincase and a poorly developed ros- sizes than males. This feature has been recorded trum, and the principal postnatal changes affected the for Pontoporia blainvillei, phocoena, Pla- rostrum and the temporal fossa, both structures tanista gangetica, bairdii, and species implied functionally to the feeding apparatus, thus sug- of the genus (Kasuya and Brow- gesting a specialized mode for catch fast prey in P. nell, 1979; Lockyer et al., 1988; Ralls and Mesnick, blainvillei. Osseous elements associated with sound 2002; Galatius, 2005). Several studies in cetaceans production were already well developed on immature address the issues of cranial ontogeny and sexual dolphins, suggesting the importance of this apparatus dimorphism (e.g., Van Waerebeek, 1993; Turner since the beginning of postnatal life. Sexual dimor- and Worthy, 2003), although only a few approach phism was detected on both shape and size variables. Females were bigger than males, in accordance with these issues through geometric morphometrics previous studies. Shape differences between sexes were (see Galatius et al., 2011; Frandsen and Galatius, found on the posterior part of premaxillaries and exter- 2013). Even though traditional morphometry nal bony nares (P < 0.01), suggesting that this sexual dimorphism is related to differences on vocalization capabilities. J. Morphol. 000:000–000, 2014. VC 2014 Wiley Additional Supporting Information may be found in the online ver- Periodicals, Inc. sion of this article.

KEY WORDS: geometric morphometrics; ontogeny; Pon- Contract grant sponsor: ANPCyT; Contract grant numbers: PICT toporia blainvillei; sexual dimorphism; skull 2008-1798 and PICT 2012-1588.

*Correspondence to: Daniela L. del Castillo; Laboratorio de Ecologıa, Comportamiento y Mamıferos Marinos, Division INTRODUCTION Mastozoologıa, Museo Argentino de Ciencias Naturales “Bernardino Cetacean skull morphology differs radically from Rivadavia”Av. Angel Gallardo 470 (C1405DJR), Buenos Aires, other mammals, as they are fully adapted to Argentina. E-mail: [email protected] aquatic life, and they can provide valuable infor- Author contributions: Study design was done by del Castillo, Flores, mation about cetacean life history (e.g., Reiden- Cappozzo. Acquisition of data was done by del Castillo. Analysis berg, 2007; Kurihara and Oda, 2009; Frandsen and interpretation of data was done by del Castillo, Flores, Cap- and Galatius, 2013). One of the main skull rear- pozzo. Manuscript preparation was done by del Castillo, Flores, rangements is the telescoping, that is, the disposi- Cappozzo. Statistical analysis was made by del Castillo. tion of cranial bones in cetaceans, including the back expansion of the maxillary bones, elongated Received 9 April 2014; Revised 2 June 2014; Accepted 5 July 2014. rostrum and dorsal position of the nares and narial passages (Miller, 1923). Telescoping forms a Published online 00 Month 2014 in frontal concavity occupied by structures involved Wiley Online Library (wileyonlinelibrary.com). in sound generation mechanisms (Ketten, 1992). DOI 10.1002/jmor.20309

VC 2014 WILEY PERIODICALS, INC. 2 D.L. DEL CASTILLO ET AL. provides important information regarding changes Geometric morphometrics allow us to study separately the two in morphology, it exhibits some limitations such as components of the form (shape and size), providing information the difficulty to separate the components of shape that is not possible to obtain through traditional morphometry (e.g., Rohlf and Marcus, 1993; Adams et al., 2004). Several authors and size (Rohlf and Marcus, 1993). Cranial studies (e.g., Zelditch et al., 2004; von Cramon-Taubadel et al., 2007; Vis- of the franciscana dolphin, P. blainvillei, has been cosi and Cardini, 2011; Gillick, 2012; Cardini, 2013), reported a addressed from qualitative and morphometric limitation associated with Procrustes-based geometric morpho- (both linear and geometric) approaches (e.g., metric analyses, known as “Pinocchio effect.” The positions of the landmarks represent shape differences as a whole but the method Flower, 1867; Pinedo, 1991; Mazzetta, 1992; Higa cannot necessarily convey the amount of shape variation occurring et al., 2002; Ramos et al., 2002; Trimble and Pra- at individual landmarks between individuals (Viscosi and Cardini, deri, 2008; Negri, 2010). Higa et al. (2002) pub- 2011). When significant variation between landmark configura- lished the only study applying a geometric tions is limited to only one or a few landmarks within the configu- morphometric approach for this species, although ration, then the variation between these landmarks may be spread across all of the landmarks used (Zelditch et al., 2004; von focused on the exploration of sexual dimorphism Cramon-Taubadel et al., 2007). This may misleadingly decrease and differences between populations in mature the variation occurring at different landmarks and, possibly, gen- specimens. erate inconsistent estimators of mean form and shape (Lele, P. blainvillei is the only extant species of the fam- 1993). In this sense, we consider such artifact regarding to the ros- ily Pontoporiidae (May-Collado and Agnarsson, tral landmarks of Pontoporia skull (see below). Superimposition of landmark configurations were performed 2006; McGowen et al., 2009), and is one of the small- by a generalized procrustes analysis (Goodall, 1991; Rohlf, est cetaceans, maximum body length reported is 1999) and size was measured as the centroid size. Collected 171 and 152 cm for females and males, respectively data was analyzed with MorphoJ (Klingenberg, 2011) and Info- (Kasuya and Brownell, 1979; Botta et al., 2010). P. stat (Di Rienzo et al., 2008). blainvillei, together with geoffrensis, form the sister group to the Delphinoidea, which includes Ontogenetic Changes and Sexual monodontids, , and ocean dolphins (Yan Dimorphism et al., 2005). The species occurs from south-eastern To analyze the main changes in shape during ontogeny, we Brazil (Siciliano, 1994) to northern Patagonia, worked with the covariance matrix of the symmetrical compo- Argentina, and inhabits coastal-marine waters, nent of the shape (Klingenberg et al., 2002). We conducted a from the coast to the 30-m isobath (Crespo et al., principal component analysis (PCA) to identify major compo- 2010). Sexual maturity is reached between 2 and 4 nents of variation of the entire sample. The rostrum is clearly the most prominent structure in P. blainvillei skull, and pre- years (Brownell, 1989; Danilewicz et al., 2000; Pan- sumably would show the major shape differences between ebianco et al., 2012) and life span is about 20 years immature and mature dolphins. Taking this into account, we (Pinedo and Hohn, 2000). performed the analyses excluding the landmarks corresponding We aim to identify major changes in shape dur- to the rostrum (i.e., landmarks 1, 22, 38, and 39, Fig. 1), thus avoiding the inconsistent mean shape estimators associated to ing postnatal ontogeny, as well as differences in the “Pinocchio effect” (see above). Finally, we performed a size- the skull shape and size associated with sex, relat- corrected PCA (i.e., PCA of the residuals of the regression of ing these changes to relevant factors in the life shape into size), to test if shape differences between sexes were history of this species, and comparing them with size-related. previously reported data on Delphinoidea. Our We performed multivariate regressions of the Procrustes goal is to contribute to the understanding of devel- coordinates against the centroid size to detect how shape varia- tion is associated with size. For comparison between males and opment of the skull of P. blainvillei, and provide females, the angles between regression vectors of both sexes new information of a family in which the knowl- were compared. Angles were computed as the arccosines of the edge is still limited. This analytical approach will signed inner products between the regression vectors (Drake expand the current knowledge about skull devel- and Klingenberg, 2008; Klingenberg and Marugan-Lob on, 2013). As ontogenetic vectors were not statically different opment of odontocete cetaceans. between sexes (the angle between the two regression vectors was 8.22; P < 0.001, under the null hypothesis that the vectors have random directions in the shape tangent space), data were MATERIALS AND METHODS pooled together and shape variation during ontogeny was ana- Sample Size and Cranial Landmarks lyzed for the entire sample. The analysis was performed using MorphoJ (Klingenberg, 2011). We analyzed a sample of 156 sexed skulls of a complete onto- To analyze shape differences between sexes, we worked with genetic series (73#,83$)ofPontoporia blainvillei (Gervais and the covariance matrix of the symmetrical component of the d’Orbigny, 1844), deposited on the collection of Museo shape, also discarding the landmarks corresponding to the ros- Nacional de Historia Natural (see Supporting Information), trum (Fig. 1). First, we run these analyses with the complete from Uruguay. Following Kasuya and Brownell (1979), we con- ontogenetic series, and then on immature and mature individu- sidered mature all males with total length of body (TL) equal als as separate subsets to detect if sexual dimorphism on shape or greater than 131 cm, and females with TL equal or greater or size was only expressed in adults periods, or dimorphism is than 140 cm. In our sample, 51 males and 59 females were con- already expressed during juvenile stages of growth. We con- sidered as immature. In five dolphins, we could not assign ducted a PCA for each subset, and when some difference on the maturity stage, because TL data were not available. We digi- shape space was detected, we proceeded to explore sex differen- tized 57 landmarks on each skull (Fig. 1) with a 3D-digitizer ces with a discriminant analysis (DA). DA was performed to VR Microscribe , which were chosen following Sydney et al. (2012) detect the landmarks showing most variation between sexes, with some modifications. Each skull was digitized at least twice because DA tends to over-estimate groups differences when the and an average of both configurations was taken. sample size is small relative to the number of landmarks

Journal of Morphology SKULL ONTOGENY OF FRANCISCANA DOLPHIN 3

Fig. 1. Pontoporia blainvillei landmarks in dorsal (A), lateral (B), and ventral (C)viewsoftheskull.

(which is our case). We select only a few landmarks recognized specimens of both sexes were placed on the posi- as the most variables between sexes on the DA (see Results) tive (right) side of the first component, although and conducted a PCA for each subset (mature and immature samples) with such selected landmarks. When difference on the few immature specimens (male and female) were shape space was detected, we also proceeded to explore sex dif- also placed on the right side of this component. ferences with a DA. Finally, we performed ANOVA analyses Mature (and some immature specimens as well) using centriod size as dependent variable, to analyze size differ- female dolphins occupied the right extreme on the ences between males and females. Here, we also carried out these analyses with the complete ontogenetic series, as well as first component, denoting the most elongated ros- on immature and mature subsets separately. We tested the nor- tral morphology, and immature and mature mor- mality of the sample with a Shapiro–Wilks test in both our phology was more clearly distinct in males (Fig. 2) complete sample and cutoff subsets (complete sample: W 5 0.99, than in females. P 5 0.86; immature subset: W 5 0.97, P 5 0.07; mature subset: W 5 0.95, P 5 0.34). Cranial shape variation excluding the rostrum is described in Figure 3. The First PC explained 16.95% of the variance, mature female dolphins RESULTS were mostly placed on the right side of this compo- Ontogenetic Changes nent, and male dolphins were more dispersed on Rostrum was the structure that showed most this component. Cranium shape was similar variation in our sample as evidenced the PCA that according to sex (except the extreme position of takes into account all cranial landmarks (Fig. 2; mature female dolphins), and some differences first PC explained 67.15% of the variance and was between mature and immature specimens were highly related to the rostrum length). Most mature detected (although with some overlapping), thus

Journal of Morphology 4 D.L. DEL CASTILLO ET AL.

Fig. 2. Pontoporia blainvillei PCA of the 3-dimensional shape variables including all land- marks. Circles: male dolphins; triangles: female dolphins. Light gray: immature dolphins; dark gray: mature dolphins. denoting a continuous pattern of variation. Sorting Sexual Dimorphism according to sexual maturity on the first PC was Sexual dimorphism was not evident using PCA more obvious in the PCA analysis that included including both mature and immature dolphins. rostral landmarks (Figs. 2 and 3). Therefore, we divided the dataset in subsets of As mentioned above, PCA analysis showed that individuals, mature and immature, and explored females have a longer allometric trajectory than sexual dimorphism in both groups. The PCA corre- males (Fig. 2), suggesting that overall shape differ- sponding to immature dolphins did not show any ences between sexes could be associated with size. special discrimination corresponding to sexual This fact was confirmed using the size corrected dimorphism, in comparison with that correspond- PCA which did not show spatial differences on the ing to mature dolphins that did show a spatial shape-space between sexes, thus indicating that pattern according to sex. The DA performed to differences on skull shape associated with sex are explore those sexual differences showed that main caused by allometric variation (Supporting Infor- differences between sexes were similar to those mation Fig. S1). observed between the extremes of the regression The regression of Procrustes coordinates versus analysis (see Fig. 4; i.e., the skull shape of female centroid size including all the 57 landmarks was highly significant (P < 0.01) and the size explained 54.1% of the shape changes. However, the same analysis excluding rostrum landmarks (Fig. 4) was also significant (P < 0.01) but size explained only 10.9% of the shape changes. Ontogenetic trajectory describes shape variation from skulls showing a rounded general shape, a not developed nuchal crest, orbits pointing to the front, rounded and thin temporal fossa, thin rostrum base pointing down, wide posterior part of premaxillary bone, mandibular (glenoid) fossa close to midline, and posterior part of the skull more developed, toward skulls with a more triangular global shape, a well- developed nuchal crest, orbits pointing more later- ally, oval and wide temporal fossa, wide rostrum base aligned with the skull, thin posterior part of Fig. 3. Pontoporia blainvillei PCA of the 3-dimensional shape premaxillary bone, mandibular fossa placed fur- variables excluding landmarks 1–22, 38–39. Circles: male dol- ther away from the midline and anterior part of phins; triangles: female dolphins. Light gray: immature dol- the skull more developed (Fig. 4). phins; dark gray: mature dolphins.

Journal of Morphology SKULL ONTOGENY OF FRANCISCANA DOLPHIN 5

Fig. 4. Pontoporia blainvillei procrustes distance from average shape plotted against CS. Trian- gles: females, Circles: males. (A, D) Dorsal; (B, E)lateral;(C, F) frontal views of lower and upper size extremes, respectively (magnification 32). dolphins is alike to the shape of the mature dol- F 5 0.37, P 5 0.54). However, centriod size showed phins, whereas the skull shape of male dolphins is sexual dimorphism for mature dolphins favoring alike to the shape of the immature dolphins). females (ANOVA, n 5 41, F 5 9.29, P < 0.01). Sex- Moreover, additional differences on the posterior ual differences in centriod size revolved the same part of the premaxillaries and external bony nares pattern when the complete landmark configuration were also detected in the DA (P < 0.01). The latter was analyzed (ANOVA complete data set: n 5 156, region showed clear differences according to sex, F 5 2.52, P 5 0.11; immature dolphins: n 5 110, evidenced on Figure 5 (PCA performed taking into F 5 0.48, P 5 0.49; mature dolphins: n 5 41, account only the landmarks implied in such area; F 5 16.63, P < 0.01). i.e., landmarks 6–17, 7–16, 8–15, 9, 10–14, 56–57), and on Table 1, that shows a high rate of correct classification with cross validation method. DISCUSSION Sexual differences in centriod size were tested Ontogenetic Changes by ANOVA excluding rostrum landmarks and no The rostrum was the feature that showed the significant differences were found (ANOVA most drastic growth during postnatal ontogeny. n 5 156, F 5 0.48, P 5 0.49). A similar outcome was This growth pattern was also observed in other found for immature dolphins (ANOVA, n 5 110, dolphins such as Tursiops spp. and

Journal of Morphology 6 D.L. DEL CASTILLO ET AL.

Fig. 5. Pontoporia blainvillei PCA of the 3-dimensional shape variables of the narial region. Only mature dolphins are plotted. Circles: male dolphins; triangles: female dolphins.

guianensis (e.g., Kurihara and Oda, 2009; Sydney (e.g., Segura and Prevosti, 2012; Flores et al., et al., 2012), although in some paedomorphic spe- 2013; Segura et al., 2013; Tarnawski et al., 2014a, cies such as porpoises and Cephalorhynchus com- 2014b). Although the longer rostrum suggests a mersonii the rostrum was not particularly decreasing of the mechanical advantages at the tip elongated during ontogeny (Galatius et al., 2011). (Preuschoft and Witzel, 2002; Segura and Prevosti, According to Werth (2006), mandibular bluntness 2012; Segura et al., 2013), its length, associated on odontocete cetaceans is related to feeding strat- with a well-developed temporal musculature, also egy, in which species with blunt heads and wide implies a gain in mouth-closing speed at the tip, jaws have more circular mouth opening, improving making prey catch function in adults more effi- water flow for suction feeding. Taking these into cient. These observations are in agreement with account, the drastic growth of the rostrum in those by Loy et al. (2011), who suggested that the adults of P. blainvillei (Fig. 2) is indicative of spe- elongation of the rostrum might be related to dif- cific feeding strategies, where suction feeding is ferences in feeding habits for different populations likely less important than a catching mode of pre- on coeruleoalba. dation. Young and adults franciscana dolphins Mature franciscana dolphins showed a rostrum exhibit differences in their feeding habits more aligned with the skull -and with the column- (Rodrıguez et al., 2002), which is also reflected in than immature dolphins. This fact has been also a notably longer rostrum in adults, probably asso- observed by Sydney et al. (2012) on Sotalia guia- ciated with a better performance of catching big- nensis, who suggested that this feature is associ- ger and faster preys. The small teeth disposed in a ated with a shift toward a more pelagic feeding. proportionally long tooth row work as a highly effi- Sound production in odontocete cetaceans is cient tool for catching preys. The shape change driven by pressurized air within a complex nasal pointed to enlarging the temporal fossa (Fig. 4) is probably associated with this performance; a lon- TABLE 1. Pontoporia blainvillei classification summary shape ger rostrum acting through well-developed tempo- variables using cross-validation ral muscles may allow dolphins to catch fast preys efficiently. In addition, the lateral displacement of Frequency of female Frequency of male the squamosal in adults (Fig. 4) derives in an Sex classification (n) classification (n) amount of the available space for the accommoda- Female 17/21 4/21 tion of the temporalis. This character seems to be Male 2/19 17/19 common in mammals (and perhaps a plesiomor- Total 19/40 21/40 phic condition), as was detected in the ontogeny of DA performed with the landmarks: 6–17, 7–16, 8–15, 9, 10–14, several groups of marine and terrestrial mammals 56–57.

Journal of Morphology SKULL ONTOGENY OF FRANCISCANA DOLPHIN 7 system (Madsen et al., 2012), where sound vibra- because neural processing needs to be associated tions propagate through the and emerge with either echolocation per se or its elaboration into the environment (McKenna et al., 2012). The into a complex perceptual system in odontocete echolocation system allows dolphins to forage, cetaceans, leading to an increase in encephaliza- avoid obstacles, and orient under water (Thomas tion at the origin of the clade (Marino et al., 2004). et al., 2004), and involves sound production as well as sound receiving. According to Cranford et al. (1996), premaxillary bones are linked to the Sexual Dimorphism shape and position of soft tissues responsible for Our results showed inverted sexual dimorphism sound production and probably influence the sonar of size, consistent with previous studies in Pontopo- beam formation process. In this sense, the early ria and other dolphins (e.g., Lahille, 1899; Kasuya development of these bones in P. blainvillei may and Brownell, 1979; Mazzetta, 1992; Pinedo, 1991; be associated with the precocious development of Higa et al., 2002). Although the inverted dimor- the sound production apparatus. Previous studies phism of Pontoporia is not a novelty, we discuss the performed on odontocetes (e.g., Hendry, 2004; Li specific shape differences between sexes. et al., 2007; Favaro et al., 2013) indicate that The main shape differences between sexes were sound emissions in some species (Tursiops trunca- associated with size differences, as noted in Figures tus and Neophocaena phocaenoides asiaeorienta- 2 and Supporting Information S1. The skull shape lis), starts at 22 days of age or earlier. Although of female dolphins is alike to the one of mature dol- there is no data about the age at which echoloca- phins, whereas the shape of the skull of male dol- tion actually begins in P. blainvillei, the early phins is alike to the one of the immature dolphins. development of such capacity is probably a charac- This fact probably occurs because females grow for teristic widespread in odontocete cetaceans (e.g., a longer period of time than males, which is in Yurick and Gaskin, 1988); the proportionally wider agreement with Barreto and Rosas (2006) and Negri premaxillary bones observed in immature francis- (2010), whom suggested that dimorphism is reached cana dolphins (Fig. 4) may be associated with that by growth extension of the trajectories of females early development. In this context, Racicot and (i.e., hypermorphosis, sensu Leigh, 1992). Moreover, Berta (2013) recently found that pterygoid sinuses an alternative explanation is that female dolphins on paedomorphic porpoises are already present on start its postnatal growth with a larger size than neonates, suggesting the importance of bidirec- males. However, our regression analysis of Pro- tional hearing in earlier stages of the ontogeny. crustes distances on CS (Fig. 4) seems to indicate Our results reinforced all previous evidence that females are not larger than males on the ear- pointed to the early acquisition of echolocation sys- lier stages of growth. Indeed, the youngest female of tem in odontocete cetaceans. our sample exhibit lesser CS than the youngest The dorsal displacement of the midpoint of the males analyzed. These results suggest that sexual nuchal crest (Fig. 4) in adults may increase the dimorphism in skull shape is size-related, as has surface for insertion of the muscles inserted on the been reported in other mammals (e.g., Cardini and occipital plate, such as the m. semispinalis (Strick- Elton, 2008; Flores and Casinos, 2011; Tarnawski ler, 1980), thus increasing the stability during et al., 2014a, 2014b), although, as a general trend, swimming. The nuchal crest is anteriorly displaced favoring males. during ontogeny, in agreement with Galatius et al. Different hypotheses have been proposed to (2010), who suggested that this displacement may explain inverted sexual dimorphism in size be associated to the supraoccipital telescoping over observed on small cetaceans. Ralls (1976) proposed the parietal and frontal bones during ontogeny. that larger body size on females exhibits better The general direction of the shape change dur- chances in competition for resources, but an alter- ing postnatal ontogeny observed on P. blainvillei native (not exclusionary) explanation pointed to skull was expected according to the common pat- the physiological needs of a minimum newborn tern observed for mammals and particularly for size to maintain body temperature. Our results of odontocetes, that is, an enlargement of the rostral size and size-related shape sexual dimorphism part of the skull, and a compression of the neuro- observed on mature dolphins are in accordance cranium, which is proportionally larger in earlier with both hypotheses. stages of postnatal development (Moore, 1966; Gal- The structure of the premaxillary bones and the atius et al., 2011; Sydney et al., 2012; Flores et al., nasal opening, condition the morphology of soft tis- 2013; Segura et al., 2013). Specific features, such sues responsible for vocalization capabilities as the early development of the premaxillaries and (Cranford et al., 1996). In this sense, sexual differ- external nares, are probably restricted to odonto- ences on this area detected in our DA may be cetes, because the skull function in this group is related to possible differences on the vocalizations not only pointed to trophic and sensitive functions, between sexes. Previously, Sayigh et al. (1995) but also to sound production mechanisms. Echolo- have suggested that on matrilines of cetaceans, cation is also believed to influence skull size vocalizations capabilities may be under greater

Journal of Morphology 8 D.L. DEL CASTILLO ET AL. selection pressures on females, as they would need kindly helped with the figures of this manuscript, to be capable to produce whistles different from and Marıa Victoria Panebianco for the English cor- their mothers to be recognized. Fripp et al. (2005) rections. This research project is a part of DdC’s suggested that males and females Tursiops may PhD thesis for which the Consejo Nacional de use signature whistles differently as adults, based Investigaciones Cientıficas y Tecnicas (CONICET) on whistle recordings of wild dolphins. Recent from Argentina granted her a postgraduate fellow- studies have suggested that matrilines would be ship. Finally, thanks to the two anonymous the social unit on franciscana dolphins (Valsecchi reviewers for their valuable corrections and sug- and Zanelatto, 2003; Costa-Urrutia et al., 2012), gestions which greatly improved this manuscript. for which females and males franciscana dolphins may have a distinct selection pressure for vocaliza- LITERATURE CITED tion capabilities, reflected on the soft tissues responsible for this, as well as in the osseous ele- Adams DC, Rohlf FJ, Slice DE. 2004. Geometric morphomet- ments supporting them. Even though whistles rics: Ten years of progress following the ‘revolution’. Ital J Zool 71:5–16. have not been recorded in franciscana dolphins Barreto AS, Rosas FC. 2006. Comparative growth analysis of (Melcon et al., 2012; Morisaka, 2012), previous two populations of Pontoporia blainvillei on the Brazilian studies proposed that the species that do not whis- coast. Mar Mamm Sci 22: 644–653. tle may compensate this loss by using other types Botta S, Muelbert MC, Secchi ER, Danilewicz D, Negri MF, of sound or clicks for communication (Morisaka, Cappozzo HL, Hohn A. 2010. Age and growth of Franciscana Dolphin, Pontoporia blainvillei, () incidentally caught 2012). Taking this into account, it is clear that fur- off southern Brazil and northern Argentina. J Mar Biol Assoc ther research is needed to elucidate the nature U. K. 90:1493–1500. and possible consequences of sexual dimorphism Brownell RL Jr. 1989. Franciscana, Pontoporia blainvillei (Ger- observed in this area. Even though the report of vais and d’Orbigny 1844). In: Ridgway SH and Harrison RJ, editors. Handbook of Marine Mammals. London: Academic Higa et al. (2002) detected absence of sexual press. pp 45–67. dimorphism on skull shape of P. blainvillei, our Cardini A. 2013. Geometric morphometrics. In: UNESCO- results indicate some shape differences. However, EOLSS Joint Committee editors. Biological Science Funda- we consider that Higa’s finding is not in complete mental and Systematics. UK: Oxford. disagreement with our results, mainly due to that Cardini A, Elton S. 2008. Variation in guenon skulls (II): Sex- ual dimorphism. J Hum Evol 54:638–647. the former authors considered only one landmark Costa-Urrutia P, Abud C, Secchi ER, Lessa EP. 2012. Popula- in the area where we detected sexual dimorphism tion genetic structure and social kin associations of francis- in shape (i.e., the nasal opening). cana dolphin, Pontoporia blainvillei. J Hered 103:92–102. Summarizing, our results indicate that the main Cranford TW, Amundin M, Norris KS. 1996. Functional mor- shape changes on the postnatal ontogeny of P. phology and homology in the odontocete nasal complex: Impli- cations for sound generation. J Morphol 228:223–285. blainvillei are associated to the changes in feeding Crespo EA, Pedraza SN, Grandi MF, Dans SL, Garaffo GV. habits, as it affects principally to the trophic appa- 2010. Abundance and distribution of endangered Franciscana ratus and its capabilities. Moreover, the early dolphins in Argentine waters and conservation implications. development of the bony elements associated with Mar Mamm Sci 26:17–35. sound production, suggests the major importance Danilewicz DS, Secchi ER, Ott PH, Moreno IM. 2000. Analysis of the age at sexual maturity and reproductive rates of fran- of this apparatus since the beginning of postnatal ciscana (Pontoporia blainvillei) from Rio Grande do Sul, life. Finally, sexual dimorphism found on the dor- southern Brazil. Comun Mus Cienc Tecnol PUCRS 13:89–98. sal part of the premaxillaries and the external Di Rienzo JA, Casanoves F, Balzarini MG, Gonzalez L, Tablada bony nares may be associated to differences on M, Robledo CW. 2008. InfoStat, Version 2008, Grupo InfoStat, vocalization capabilities according to sex, a topic FCA, Argentina: Universidad Nacional de Cordoba. Drake AG, Klingenberg CP. 2008. The pace of morphological that should be also addressed in a behavioral con- change: historical transformation of skull shape in St Ber- text in this species. The detection of additional nard dogs. Proc R Soc B: Biol Sci 275:71–76. sexual differences and its expression during the Favaro L, Gnone G, Pessani D. 2013. Postnatal development of ontogeny in other delphinid species is also highly echolocation abilities in a (Tursiops trunca- interesting and necessary to detect patterns of tus): Temporal organization. Zoo Biol 32:210–215. Flores DA, Casinos A. 2011. Cranial ontogeny and sexual dimorphism and ontogenetic shape changes in dimorphism in two new world monkeys: Alouatta caraya (Ate- phylogenetic and functional grounds, considering lidae) and Cebus apella (Cebidae). J Morphol 272:744–757. the strongly specialized life history in cetaceans, Flores DA, Abdala F, Giannini NP. 2013. Post-weaning cranial and the different habitats (i.e., coastal and off- ontogeny in two bandicoots (Mammalia, Peramelomorphia, Peramelidae) and comparison with carnivorous marsupials. shore species) where they inhabits. Zoology 116:372–384. Flower WH. 1867. IV. Description of the skeleton of Inia geof- frensis and of the skull of Pontoporia blainvillii, with ACKNOWLEDGMENTS remarks on the systematic position of these in the The authors thank Enrique Gonzalez for permit- order Cetacea. Trans Zool Soc London 6:87–116. Frandsen MS, Galatius A. 2013. Sexual dimorphism of Dall’s ting access to collection specimens and Sabrina and harbor porpoise skulls. Mamm Biol 78:153–156. Riveron who was very hospitable during sampling. Fripp D, Owen C, Quintana-Rizzo E, Shapiro A, Buckstaff K, The authors also thank Marcelo del Castillo who Jankowski K, Wells R, Tyack P. 2005. Bottlenose dolphin

Journal of Morphology SKULL ONTOGENY OF FRANCISCANA DOLPHIN 9 (Tursiops truncatus) calves appear to model their signature May-Collado L, Agnarsson I. 2006. Cytochrome b and Bayesian whistles on the signature whistles of community members. inference of phylogeny. Mol Phylogenet Evol 38:344– Anim Cogn 8:17–26. 354. Galatius A. 2005. Sexually dimorphic proportions of the har- Mazzetta GV. 1992. Alometrıa de tamano~ en la especie Pontopo- bour porpoise (Phocoena phocoena) skeleton. J Anat 206:141– ria blainvillei (Cetacea: Platanistoidea). Anales III Reunion 154. de Trabajo de Especialistas en Mamıferos Acuaticos de Galatius A. 2010. Paedomorphosis in two small species of America del Sur: 29–35. toothed (Odontoceti): How and why? Biol J Linn Soc McGowen MR, Spaulding M, Gatesy J. 2009. Divergence date 99:278–295. estimation and a comprehensive molecular tree of extant Galatius A, Berta A, Frandsen MS, Goodall RNP. 2011. Inter- cetaceans. Mol Phylogenet Evol 53:891–906. specific variation of ontogeny and skull shape among por- McKenna MF, Cranford TW, Berta A, Pyenson ND. 2012. Mor- poises (Phocoenidae). J Morphol 272: 136–148. phology of the odontocete melon and its implications for Gillick H. 2012. Ancestry determination using geometric mor- acoustic function. Mar Mamm Sci 28:690–713. phometrics. [Master of Science Thesis]. Scotland, United Melcon ML, Failla M, In~ıguez MA. 2012. Echolocation behavior Kingdom: University of Dundee. 122p. of franciscana dolphins (Pontoporia blainvillei) in the wild. J Acous Soc Am 131:EL448–EL453. Goodall C. 1991. Procrustes methods in the statistical analysis Miller GS. 1923. The Telescoping of the Cetacean Skull. Smith of shape. J R Stat Soc Series B Stat Methodol 53:285–339. Misc Coll 76:1–70. Hendry JL. 2004. The ontogeny of echolocation in the Atlantic Moore WJ. 1966. Skull growth in the albino rat (Rattus norve- bottlenose dolphin (Tursiops truncatus) [Ph.D. dissertation] gicus). J Zool 149:137–144. Mississippi, USA: The University of Southern Mississippi. Morisaka T. 2012. Evolution of communication sounds in odon- 133p. tocetes: A review. Int J Comp Psychol 25: 1–20. Higa A, Hingst-Zaher E, Vivo MD. 2002. Size and shape vari- Negri MF. 2010. Estudio de la biologıa y ecologıa del delfın ability in the skull of Pontoporia blainvillei (Cetacea: Ponto- franciscana, Pontoporia blainvillei, y su interaccion con la poriidae) from the Brazilian coast. Lat Am J Aquat Mamm 1: pesquerıa costera en la provincia de Buenos Aires. [Ph.D. dis- 145–152. sertation]. Buenos Aires, Argentina: University of Buenos Kasuya T, Brownell RL Jr. 1979. Age determination, reproduc- Aires. 155 p. tion, and growth of the franciscana dolphin, Pontoporia blain- Panebianco MV, Negri MF, Cappozzo HL. 2012. Reproductive villei. Sci Rep Whales Res Inst 31:45–67. aspects of male franciscana dolphins (Pontoporia blainvillei) Ketten DR. 1992. The marine mammal ear: Specializations for off Argentina. Anim Reprod Sci 131:41–48. aquatic audition and echolocation. In: Webster DB, Ray RR, Pinedo MC. 1991. Development and variation of the francis- Popper AN, editors. The Evolutionary Biology of Hearing. cana, Pontoporia blainvillei.[Ph.D. dissertation]. California, New York: Springer. pp 717–750. USA: University of California. 405 p. Klingenberg CP. 2011. MorphoJ: An integrated software pack- Pinedo MC, Hohn A. 2000. Growth layer patterns in teeth from age for geometric morphometrics. Mol Ecol Resour 11:353– the franciscana, Pontoporia blainvillei: developing a model 357. for precision in age estimation. Mar Mamm Sci 16:1–27. Klingenberg CP, Barluenga M, Meyer A. 2002. Shape analysis Preuschoft H, Witzel U. 2002. Biomechanical investigations on of symmetric structures: quantifying variation among individ- the skulls of reptiles and mammals. Senckenb Lethaea 82: uals and asymmetry. Evolution 56: 1909–1920. 207–222. Klingenberg CP, Marugan-Lob on J. 2013. Evolutionary covaria- Racicot RA, Berta A. 2013. Comparative morphology of porpoise tion in geometric morphometric data: analyzing integration, (Cetacea: Phocoenidae) pterygoid sinuses: Phylogenetic and modularity, and allometry in a phylogenetic context. Syst Biol functional implications. J Morphol 274:49–62. 62:591–610. Ralls K. 1976. Mammals in which females are larger than Kurihara N, Oda S. 2009. Effects of size on the skull shape of males. Q Rev Biol 51:245–276. the bottlenose dolphin (Tursiops truncatus). Mamm Study 34: Ralls K. 1977. Sexual dimorphism in mammals: avian models 19–32. and unanswered questions. Am Nat 111:917–938. Lahille F. 1899. Notes sur les dimensions du Stenodelphis Ralls K, Mesnick SL. 2002. Sexual dimorphism. In: Perrin WF, blainvillei. Revista del Museo de La Plata 9:389–392. Wursig€ B, Thewissen JGM, editors. Encyclopedia of marine mammals. San Diego: Academic Press. p 1071–1078. Leigh SR. 1992. Patterns of variation in the ontogeny of pri- mate body size dimorphism. J Hum Evol 23:27–50. Ramos RM, Di Beneditto APM, Siciliano S, Santos MC, Zerbini AN, Bertozzi C, Lima NR. 2002. Morphology of the francis- Lele S. 1993. Euclidean distance matrix analysis (EDMA): esti- cana (Pontoporia blainvillei) off Southeastern Brazil: sexual mation of mean form and mean form difference. Math Geol dimorphism, growth and geographic variation. Lat Am J 25:573–602. Aquat Mamm 1:129–144. Li S, Wang D, Wang K, Xiao J, Akamatsu T. 2007. The ontog- Reidenberg JS. 2007. Anatomical adaptations of aquatic mam- eny of echolocation in a Yangtze finless porpoise (Neopho- mals. Anat Rec 290:507–513. caena phocaenoides asiaeorientalis). J Acous Soc Am 122: Rodrıguez D, Rivero L, Bastida R. 2002. Feeding ecology of the 715–718. franciscana (Pontoporia blainvillei) in marine and estuarine Lockyer C, Goodall RNP, Galeazzi AR. 1988. Age and body waters of Argentina. Lat Am J Aquat Mamm 1:77–94. length characteristics of Cephalorhynchus commersonii from Rohlf FJ. 1999. Shape statistics: Procrustes superimpositions incidentally-caught specimens off Tierra del Fuego. Rep Int and tangent spaces. J Classif 16:197–223. Whal Comm (Special issue 9):103–118. Rohlf FJ, Marcus LF. 1993. A revolution morphometrics. Trends Loy A, Tamburelli A, Carlini R, Slice DE. 2011. Craniometric Ecol Evol 8:129–132. variation of some Mediterranean and Atlantic populations of Sayigh LS, Tyack PL, Wells RS, Scott MD, Irvine AB. 1995. Stenella coeruleoalba (Mammalia, Delphinidae): A three- Sex difference in signature whistle production of free-ranging dimensional geometric morphometric analysis. Mar Mamm bottlenose dolphins, Tursiops truncates. Behav Ecol Sociobiol Sci 27:E65–E78. 36:171–177. Madsen PT, Jensen FH, Carder D, Ridgway S. 2012. Dolphin Segura V, Prevosti F. 2012. A quantitative approach to the cra- whistles: a functional misnomer revealed by heliox breathing. nial ontogeny of Lycalopex culpaeus (Carnivora: Canidae). Biol Lett 8:211–213. Zoomorphology 131:79–92. Marino L, McShea DW, Uhen MD. 2004. Origin and evolution Segura V, Prevosti F, Cassini G. 2013. Cranial ontogeny in the of large brains in toothed whales. Anat Rec A: Discov Mol Puma lineage, Puma concolor, Herpailurus yagouaroundi, Cell Evol Biol 281:1247–1255. and Acinonyx jubatus (Carnivora: Felidae): a three-

Journal of Morphology 10 D.L. DEL CASTILLO ET AL. dimensional geometric morphometric approach. Zool J Linn Turner JP, Worthy GA. 2003. Skull morphometry of bottlenose Soc 169:235–250. dolphins (Tursiops truncatus) from the Gulf of Mexico. Siciliano S. 1994. Review of small cetaceans and fishery inter- J Mamm 84:665–672. actions in coastal waters of Brazil. Rep Int Whal Comm 15: Valsecchi E, Zanelatto RC. 2003. Molecular analysis of the 241–250. social and population structure of the franciscana (Pontoporia Strickler TL. 1980. The axial musculature of Pontoporia blain- blainvillei): conservation implications. J Cetac Res Manag 5: villei, with comments on the organization of this system and 69–76. its effect on fluke-stroke dynamics in the cetacea. Am J Anat Van Waerebeek K. 1993. Geographic variation and sexual 157:49–59. dimorphism in the skull of the , Lagenorhyn- Sydney NV, Machado FS, Hingst-Zaher E. 2012. Timing of onto- chus obscurus (Gray, 1828). Fish Bull 91:754–774. genetic changes of two cranial regions in Sotalia guianensis Viscosi V, Cardini A. 2011. Leaf morphology, and geo- (Delphinidae). Mamm Biol 77:397–403. metric morphometrics: a simplified protocol for beginners. Tarnawski BA, Cassini GH, Flores DA. 2014a. Allometry of the PLoS One 6:e25630. postnatal cranial ontogeny and sexual dimorphism in Otaria von Cramon-Taubadel N, Frazier BC, and Lahr MM. 2007. The byronia (Otariidae). Acta Theriol 59:81–97. problem of assessing landmark error in geometric morpho- Tarnawski BA, Cassini GH, Flores DA. 2014b. Skull allometry metrics: theory, methods, and modifications. Am J Phys and sexual dimorphism in the ontogeny of the southern ele- Anthrop 134:24–35. phant seal (Mirounga leonina). Can J Zool 92:19–31. Werth AJ. 2006. Odontocete suction feeding: experimental anal- Thomas JA, Moss CF, Vater M. 2004. Echolocation in Bats and ysis of water flow and head shape. J Morphol 267:1415–1428. Dolphins. University of Chicago Press: Chicago. 607 p. Yan J, Zhou K, Yang G. 2005. Molecular phylogenetics of ‘river Tolley KA, Read AJ, Wells RS, Urian KW, Scott MD, Irvine AB, dolphins’ and the mitochondrial genome. Mol Phylogenet Hohn A. 1995. Sexual dimorphism in wild bottlenose dolphins Evol 37:743–750. (Tursiops truncatus) from Sarasota, Florida. J Mammal 76: Yurick DB, Gaskin DE. 1988. Asymmetry in the skull of the 1190–1198. Phocoena phocoena (L.) and its relationship Trimble M, Praderi R. 2008. Assessment of nonmetric skull to sound production and echolocation. Can J Zool 66:399–402. characters of the Franciscana (Pontoporia blainvillei)in Zelditch ML, Swiderski DL, Sheets HD, Fink L. 2004. Geomet- determining population differences. Aquat Mamm 34:338– ric Morphometrics for Biologists: A Primer. Elsevier: San 348. Diego. 478 p.

Journal of Morphology